Research map / Image generation / Editing and style transfer
MSCap: Multi-Style Image Captioning With Unpaired Stylized Text
Longteng Guo, Jing Liu, Peng Yao, Jiangwei Li, Hanqing Lu
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add adversarial loss (ablation) | MSCap · Four styles average | CIDEr: 14.6 -> 33.1 | Improved |
| Add adversarial loss (ablation) | MSCap · Four styles average | perplexity: 57.1 -> 20.5 | Improved |
| Add style classification loss (ablation) | MSCap · Four styles average | style accuracy: 30.0 -> 91.5 | Improved |
| Add style classification loss (ablation) | MSCap · Four styles average | CIDEr: 20.5 -> 33.1 | Improved |
| Add back-translation loss (ablation) | MSCap · Four styles average | CIDEr: 7.72 -> 33.1 | Improved |
| Add back-translation loss (ablation) | MSCap · Four styles average | perplexity: 13.6 -> 20.5 | Hurt |
| Add back-translation loss (ablation) | MSCap · Four styles average | style accuracy: 96.0 -> 91.5 | Hurt |
| Pre-train the generator with unpaired stylized text (ablation) | MSCap · Four styles average | CIDEr: 30.3 -> 33.1 | Improved |
| Pre-train the generator with unpaired stylized text (ablation) | MSCap · Four styles average | perplexity: 22.2 -> 20.5 | Improved |
| Pre-train the generator with unpaired stylized text (ablation) | MSCap · Four styles average | style accuracy: 88.7 -> 91.5 | Improved |
Closest papers by meaning
- MemCap: Memorizing Style Knowledge for Image Captioning (AAAI 2020 · similarity 0.89)
- ``Factual'' or ``Emotional'': Stylized Image Captioning with Adaptive Learning and Attention (ECCV 2018 · similarity 0.88)
- StyleNet: Generating Attractive Visual Captions With Styles (CVPR 2017 · similarity 0.88)
- SemStyle: Learning to Generate Stylised Image Captions Using Unaligned Text (CVPR 2018 · similarity 0.88)
- Controllable Image Captioning via Prompting (AAAI 2023 · similarity 0.87)
- Adversarial Semantic Alignment for Improved Image Captions (CVPR 2019 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
